Session/Mixdown practices.

[mrcbass]

I do a bit of amateur recording, typically for demos for a band I'm in. I'm not a very experienced engineer (don't do it enough) but can produce reasonable results, especially considering i use about 20% of the tool's features. This question isn't a technical question so it shouldn't matter, but I use the free version of Presonus Studio One for my DAW.

I'm going through finishing up demos right now and had a request to revert back to a previous version and then perform edits and generate another version. I had not preserved the DAW, so I really couldn't easily "revert". I had the previous mixdown and was able to manually "turn back the clock", but the edits were minor. I would have been screwed if the edits that I needed to undo were extensive.

Question regarding your archival practices: Do you freeze your DAW session after exporting a mixdown? By "freeze" I'd presume doing a "save as" version number after each mixdown edit and then creating a new session (based on the latest) for each subsequent version.

It seems like a good idea and a little anal all at the same time.

 

[Bob_Ross]

It seems like a good idea and a little anal all at the same time.

QFT

 

[LBS-bass]

I do archive my "good" versions, and by that I mean the ones I'm happy enough with to want to listen to on multiple systems. Sometimes I want to go back to someplace I was before and start over if I haven't got things quite right, so this gives me the option of various starting points for that.

 

[mrcbass]

QFT

?

 

[smeet]

Not anal, just sensible. Save a numbered copy of the session each time you do a mix that you may want to come back to. Disk space is cheap, your time is not.

 

[Bob_Ross]

?

Sorry, internet shorthand. QFT = Quoted For Truth ...it's used to reinforce the notion that the cited quote is indeed correct and/or endorsed by someone else. Kinda like saying "+1"

 

[teleharmonium]

I will sometimes save an important revision before moving on with the working copy of the project. No way could I do that with every change.

Sometimes not being able to do something for basic technical reasons is good.

 

[And I]

first thing i do when i open up a session file to start working on it is do a "save as". there is literally no reason not to save every iteration of your session files. session files from most DAWs are very small files.

 

[silky smoove]

I do this professionally. My standard practice is to do all of the tracking in one session that remains unedited. I save a copy and do all of the editing, comping, tuning, etc. that needs to occur prior to mixing as a new session. When that's done I save a new session that's specific for mixing but don't begin mixing until the edit session has been approved by the client. Beyond that I don't save new sessions for updated mixes. If someone wants me to go back to a previous mix I listen to that mix and work to recapture some or all of the vibe that it had. It's probably not the most efficient way to work, but it's always worked for me. If someone wants a prior mix I ask what it is about that earlier version that they want brought back as it's almost never the entire mix but rather a single element (i.e. "the kick hit a lot harder on the 2022/03/02 version of the mix," for instance).

 

[smeet]

first thing i do when i open up a session file to start working on it is do a "save as". there is literally no reason not to save every iteration of your session files. session files from most DAWs are very small files.

That's my feeling. I don't do this professionally, but I work with animators, who have a similar workflow. By the time they publish a scene (analogous to creating an audio mix), they may have saved several hundred versions of the scene. You may not need them, but there's no downside, and a huge upside if you ever need to go back to something you did yesterday or last week.

 

[silky smoove]

If someone wants a prior mix I ask what it is about that earlier version that they want brought back as it's almost never the entire mix but rather a single element (i.e. "the kick hit a lot harder on the 2022/03/02 version of the mix," for instance).

Quoting myself here as a discussion point: Before you restore a session to a prior mix be sure and ask the client what it is they're trying to bring back and also what it is about the current mix that they like more than the earlier version. You can end up chasing your tail by restoring an old mix only to have them then request changes to it that are more in line with the more recent version. Bruce Swedien would often respond to mix comments by zeroing out all of the faders on his desk and starting over because according to him, if you receive a comment to "make the bass louder", for instance, you couldn't just bring up that fader a little more as there are consequence when everything is coming out of the same two speakers. Similarly, a change to better reflect an aspect of an earlier session rarely, if ever, exists in a vacuum. Change the kick and the bass feels different. Change the guitar and the keys or the BVs feel different. Mix reversion, unless the previous mix was absolutely perfect for the client and is indeed a wholesale reversion, basically never works as far as I'm concerned and in my professional experience as a mix engineer. Don't consider iterative save as operations to be a silver bullet except in the rarest of instances, but rather as a tool that might help you or might not.

 

[And I]

i wasn't suggesting it's any sort of silver bullet, but there really is no reason not to do it. worst case you never need it and you've used up a couple megabytes of disk space. i wouldn't often use it as a restore point, but as a point of reference. when i did do this work professionally i don't recall a scenario where a client had me revert to a previous session file. but i did use them from time to time to do things like an A/B comparison, or to copy a region from an old session that hadn't had any editing or comping done on it.

 

[silky smoove]

i wasn't suggesting it's any sort of silver bullet

I was speaking about the practice in general, not anything anyone in this thread has stated specifically.

but there really is no reason not to do it

For my part as a currently working professional in the field I'd add that in addition to there not being a reason not to do it, there's also not a reason in most situation to do it. It's not either or, but rather both. In other words, I don't see an appreciable downside but I also don't see an upside either. I use a platform called Mixup to distribute mixes to clients. If I want to review an old mix or recall a portion of it I'll just do it on that platform and make adjustment in my DAW on the current session. Duplicating old settings from a previous session won't sound the same in the current session, is time consuming, takes up an admittedly minimal quantity of hard drive space, and ultimately doesn't offer anything to my workflow. That's why I personally/professionally don't bother.

 

[teleharmonium]

i wasn't suggesting it's any sort of silver bullet, but there really is no reason not to do it.

For me, there is one reason... I might then have to go through all that stuff and figure out exactly what the difference is for the one I am looking for ! IOW for it to be practical it would need to be accompanied by notes which would add time steadily to the project, while (at least in my experience) being needed rarely.

The larger topic here being the relationship between the person doing the mixing and everyone else with some sort of stake in the project. If you're being paid by the hour, and/or have the latitude to work in a patient and planned way, I can see it making sense to do careful documentation along the way and lots of backups. That's never been my world, but I can see the appeal.

 

[Bob_Ross]

This may already be implicit from all the previous responses in this thread, but just in case:

Saving earlier versions of a mix -- having a "revert to" timestamp in the DAW that you can instantly call up and reissue (or use as a new start point for more revisions) ...is a good strategy for any mixes that you have issued to the client. If someone else has heard it, chances are good they may want to return to that state.

But it's completely nuts for any/all mixes that no one else but you have heard.

For every final [sic] mix I submit to the client, there can easily be 3, 4, a half dozen other mixes that I consider "done" ...and then I sleep on it, come back with fresh ears, and decide for myself which one is really getting submitted. As soon as I send one off, the other versions get archived (with important notes embedded, like "vox up 2db" or "dry drum mix") and as soon as I get feedback from the client on the issued version, I throw out those archived versions. (Unless they specifically ask for something in my archives -- "sounds great, but can I get one with the lead vocal a bit louder?" or "sounds great, but can I get one with no ambience on the drum kit?") Because I don't need the distraction/option-anxiety of all those choices; I've already made my decision.